Synthesis of Chiral Saturated Heterocycles Bearing Quaternary Centers via Enantioselective β-C(sp³)−H Activation of Lactams

Abstract

The development of catalytic methods for the synthesis of enantiopure saturated heterocycles is a long-standing challenge in asymmetric catalysis. We describe the first highly enantioselective palladium-catalyzed β-C(sp³)−H arylation and olefination of lactams for the preparation of various chiral N-heterocycles bearing quaternary carbon centers. The presence of strongly electron-withdrawing groups on the chiral bifunctional MPAThio ligand is crucial for the reactivity of weakly coordinating lactams. The resulting enantioenriched lactams are readily converted to a family of chiral piperidines and imides that are highly desirable in drug discovery.

Graphical Abstract

Chiral nitrogen-containing heterocycles are ubiquitous in bioactive natural products, pharmaceuticals and materials science.¹ Stereoselective construction of chiral lactams and saturated C3-chiral heterocycles is instrumental in the syntheses of various natural products and drug molecules (Scheme 1a).² Historically, the synthesis of C3-chiral lactams bearing a-quaternary centers has employed enolate alkylation and Michael additions, with the reaction stereoselectivity relying on chiral auxiliary chemistry.³ Catalytic examples are scarce, and transformations are driven predominantly by the process of enolate stabilization.⁴ In this context, the Stoltz group⁵ has systematically developed a number of elegant Pd(0)-catalyzed asymmetric enolate alkylation, arylation, and olefination reactions of lactams (Scheme 1b). Typically, these protocols involve the use of strong bases which may not be compatible with certain functional groups.

Scheme 1.

Asymmetric C-H activation of lactams

Since 2008, a wide range of Pd(II)-catalyzed asymmetric C−H activation reactions have been developed through desymmetrization, kinetic resolution, as well as enantioselection of methylene C−H bonds on the same carbon.⁶ Our group has reported the enantio- and diastereoselective C(sp3)−H arylation of cyclopentane, cyclohexane, and cycloheptane derivatives, along with isopropyl and linear carboxylic acid derivatives. These reactions are facilitated by acidic CONHArF (ArF = p-CF₃C₆F₄) which can coordinate with Pd(II) effectively via imidate form.^6b,7 However, neutral amide-directed enantioselective C−H activation remains elusive.⁸ Especially, lactam-directed C(sp³)-H bonds are challenging as the β-C-H bonds and the directing atom are out-of-plane (Scheme 1b).⁹ Considering the importance of chiral lactams, we began to develop chiral ligands to enable desymmetrizing C−H activation of lactam substrates containing α-gem-dimethyls. If successful, an array of lactams bearing chiral quaternary carbon centers could be prepared. Here we report a Pd(II)-catalyzed asymmetric C(sp³)−H bond arylation and olefination of 6, 7, 8, and 9-membered lactams (Scheme 1c). A wide range of chiral lactams bearing quaternary centers are obtained with high yields and enantioselectivities using (hetero)aryl iodides and olefins as coupling partners. These chiral lactams can also be reduced to chiral piperidines or oxidized to imides, both of which are widely utilized sp³ rich heterocycles in drug discovery.^1a,10

We began our investigation using 1,3,3-trimethylpiperidin-2-one (1a) as the model lactam and methyl 4-iodobenzoate (2a) as the coupling partner to identify suitable chiral ligands and optimize reaction conditions (Table 1). A series of bifunctional chiral ligands previously reported for promoting C(sp³)–H activation, including MPAA (L1),¹¹ mono-N-protected aminomethyl oxazoline (MPAO) (L2),^9a mono-benzoyl protected amino thioether ligand (L3),¹² and N-acyl sulfonamide ligands (L4)¹³ failed to provide the desired arylated product, owing to the poor reactivity of the weakly coordinating lactams. In contrast, the chiral mono-protected aminoethyl phenyl thioether (MPAThio) ligand (L5)¹⁴ afforded the desired arylation product 3a with high enantioselectivity, albeit with low activity. Initial attempts to improve the yield by adding alkyl substituents to the meta-position of the thiophenol moiety (L6 and L7) were unsuccessful. We hypothesized that the strong coordination between the thioether ligand and the palladium catalyst might be hindering the catalyst’s interaction with the substrate. To address this, we modified the MPAThio ligand by introducing electron-withdrawing groups into the thiophenol. This strategy proved effective, as the ligand bearing a 3,5-diCF₃-substituted thiophenol (L8) significantly improved catalyst activity, providing the desired product in 52% yield with >99% ee. Introducing sterics at the backbone of the thioether ligand did not lead to improvements (L9-L12). Finally, extending the reaction time to 56 hours increased the reaction yield to 80% with >99% ee.

Table 1.

Reaction condition screening for enantioselective arylation^a,b

^aConditions: 1a (0.10 mmol), 2a (2.5 equiv.), Pd(OAc)₂ (10 mol%), ligand (20 mol%), Ag₂O (2.0 equiv.), HFIP (0.50 mL), 50 °C, under air, 36 h.

^bYields were determined by ¹H NMR analysis of the crude reaction mixture using CH₂ Br₂ as an internal standard. The ee values were determined by chiral SFC.

^ct = 56 h.

^dIsolated yield.

With the optimized conditions established, we investigated the substrate scope by examining a range of substituted aryl and (hetero)aryl iodides (Scheme 2). Aryl iodides bearing electron-withdrawing groups and halide at para-position (2a-g) were smoothly converted to products with good yields and excellent enantioselectivities (93->99% ee). Electron-donating groups were also tolerated, producing the desired arylation products (3h, 3i) in moderate yields (30–43%) and ee (94–97%). Further exploration of meta- (2j-l) and ortho- (2m, 2n) substitutions on aryl iodides yielded chiral products in moderate to good yields (52–71%) with high enantiomeric ratios (97–99% ee). Additionally, substrates containing heterocyclic motifs—such as CF₃-substituted pyridines (2o-q), thiophene (2r), and benzofuranyl (2s)—were well tolerated, yielding corresponding products with 97 to >99% ee. Next, we turned our attention to lactams with varying functional groups and ring sizes. Modifying the N-Me group was feasible, resulting in lactams with removable functional groups such as N-Bn (3t) and N-PMB (3u), both exhibiting high enantioselectivity (97% and 98% ee). The kinetic resolution of mono-α-methyl-substituted lactams with different ring sizes (1v-x) were also converted to the desired products with high enantioselectivity (88–91% ee). The 5-methyl δ-lactam (1y), quinolin-2-one (1z), and seven-membered lactam (1aa) were successfully converted to the products with high enantioselectivities (83–92% ee), albeit with more modest yield (28–58%). Moreover, linear native amides (1ab-ad) were also exemplified to show high yield (68–80%) and enantiomeric ratios (95->99% ee). To demonstrate further impact, structurally complex aryl iodides derived from biorelevant molecules and pharmaceuticals were tested to afford 3ae-ah in 59‒79% yields with excellent stereocontrol.

Scheme 2. The substrate scope of enantioselective arylation^a,b.

Reaction conditions A: Substrate lactam 1 (0.10 mmol), (hetero)aryl iodide 2 (0.25 mmol), Pd(OAc)₂ (10 mol%), L8 (20 mol%), Ag₂O (2.0 equiv.), HFIP (0.50 mL), 50 °C, air, 56 hours. ^a 15 mol% Pd(OAc)₂ and 30 mol% L8. ^b L10 instead of L8.

Further extending this method of enantioselective C−H activation for C(sp³)−H olefination with an olefin coupling partner will open avenues to a wider array of chiral N-heterocycles bearing an alkenyl group.

Consequently, we proceeded to explore the enantioselective reactions of lactams with various mono- or disubstituted olefin coupling partners, an investigation hitherto unreported. Asymmetric olefination of lactam 1a, facilitated by chiral ligand L5, provided the product with 32% yield and 79% ee (Table 2). A comprehensive evaluation of various conditions revealed that utilizing the CF₃-substituted MPAThio ligand L12 with 10 mol % of Pd(PhCN)₂Cl₂, resulted in completion of the reaction in 56 hours at 110 °C, yielding 75% of 5a with 94% ee (Table 2).

Table 2.

Reaction condition screening for enantioselective olefination^a,b

^aConditions: 1a (0.10 mmol), 4a (3.0 equiv.), Pd(PhCN)₂Cl₂ (10 mol%), ligand (20 mol%), AgNO₃ (2.0 equiv.), HFIP (0.50 mL), 110 °C, under air, 56 h.

^bYields were determined by ¹H NMR analysis of the crude reaction mixture using CH₂ Br₂ as an internal standard. The ee values were determined by chiral SFC.

^cIsolated yield.

With the conditions for olefination optimized, the potential range of olefins and lactams was explored, as illustrated in Scheme 3. Commercially mono-substituted olefins, comprising acrylates (4a-f), vinyl sulfonates (4g, h), ethenesulfonyl fluoride¹⁵ (ESF, 4i), vinyl phosphonate (4j), and vinyl amide (4k) were smoothly converted into the target products, boasting moderate to high yields and 83–92% ee. It is worth noting that perfluorinated styrene and aliphatic alkenes were also tolerated, leading to the production of fluorinated chiral lactams (5l, m) in low to moderate yields with good enantioselectivities (80–86% ee). Further, we evaluated the potential of 1,1-disubstituted and internal olefins coupling partners. Olefination of the model lactam 1a with dimethyl itaconate (4n), methyl methacrylate (4o) and phenyl methacrylate (4p) were efficient, yielding target products with 41–81% yields and 80–90% ee. Interestingly, the internal olefin maleimide could also be successfully converted to the desired product 5q, albeit in a moderate yield, but with good enantioselectivity (80% ee). Lastly, our study also encompassed the scope of lactams possessing diverse substituents and ring sizes. Similar to arylation reaction, modifying the protecting group of lactam from N-Me to N-PMB and N-Bn was well tolerated, providing chiral lactams (5r-t) in good yields (70–82%) and enantiomeric ratios (87–91% ee). Lactams featuring either a seven-membered ring or a quinolin-2-one were subject to the olefination condition and successfully yielded the desired products (5u, 5v) in moderate to good yields and good enantioselectivities (80–90% ee). Besides, linear native amides were also converted to olefination products (5w, x) in high yields (72–82%) with excellent enantiomeric ratios (83–97% ee). Finally, the acrylic acid ester of L-menthol was smoothly converted to the chiral lactam 5y with moderate yield and good diastereoselectivity.

Scheme 3. The substrate scope of enantioselective olefination.

Reaction conditions B: Substrate lactam 1 (0.10 mmol), acrylate 4 (0.30 mmol), Pd(PhCN)₂Cl₂ (10 mol%), L12 (20 mol%), AgNO₃ (2.0 equiv.), HFIP (0.50 mL), 110 °C, 56 hours. ^a15 mol% Pd(PhCN)₂Cl₂ and 30 mol% L12. ^bL8 instead of L12.

To explore the potential utility of this method, arylation and olefination reactions were performed at 1 mmol scale. The desired arylated and olefinated products were obtained in good yields with excellent enantioselectivities (Scheme 4a). In addition, the gram-scale arylation reaction was successfully conducted, affording 0.9015 g of product 3a in 65% isolated yield. Subsequent synthetic transformations were investigated, as shown in Schemes 4b and 4c. Treatment of the arylation product 3a with CrO₃ in acetic acid yielded imide 6a in 52% yield without any loss of enantiomeric purity, along with a side product 6b¹⁶ (14% yield). Similarly, treatment 5a with CrO₃ produced imide 6c in 60% yield with 88% ee. These imide products, upon deprotection, are important cereblon binding motifs in drug discovery.¹⁷ An allylic oxidation reaction of 5a in 1,4-dioxane under reflux conditions afforded 6d in 65% yield and 91% ee. Finally, reduction of the chiral lactams with LiAlH₄ generated a variety of chiral piperidines (7a-f) featuring quaternary carbon centers (Scheme 4c). These novel chiral piperidines hold significant potential for future applications in synthetic and medicinal chemistry.

Scheme 4. Scale-up synthesis and synthetic Application.

^a Conditions: 3a or 5a (0.10 mmol), CrO₃ (3.0 equiv.), AcOH (0.1 M), 80 °C, 24 h. ^bConditions: 5a (0.10 mmol), SeO₂ (2.0 equiv.), 1,4-dioxane (0.1 M), 110 °C, 24 h. ^cConditions: 3 or 5 (0.10 mmol), LiAlH₄ (3.0 equiv.), dry THF (0.1 M), 50 °C, 3 h. ^d LiAlH₄ (6.0 equiv.).

In summary, we have achieved the first palladium-catalyzed asymmetric C(sp³)−H bond activation of lactams. This method demonstrates that the desymmetrization of the dimethyl groups in lactams can be achieved using the electron-deficient chiral MPAThio ligand. Given the ready availability of lactam substrates, (hetero)aryl iodide/olefin coupling partners, and chiral catalyst, this approach offers practical and convenient access to a broad range of enantioenriched chiral lactams. Key features of this method include excellent enantioselectivity, exceptional functional group tolerability, and high step-economy. Notably, the obtained products can be easily transformed into a series of chiral piperidines and imides. We believe this approach will provide a more convenient pathway for the synthesis of drugs and materials containing chiral sp³ rich N-heterocycles.

Supplementary Material

The Supporting Information is available free of charge on the ACS Publications website. Experimental details, full characterization of new compounds including ¹H and ¹³C NMR spectra, HRMS data (PDF)